[Arzela]

Hi all,

I want to bypass the rectifier bride in my
power supply (decouple?). The rect's are 100V/2 Amp,
feed by aproximately 22V RMS AC. Can anyone
tell me if the following caps are
suitable?

http://www.wima.com/fkp2.htm

I was thinking of values around 500 nf.

Any suggestions for more suitable caps
are welcome.

 

[Squalish]

I'm leaning towards using the end result of this article.

The only thing I've heard against using those as snubbers is someone claiming that spikes in the AC damage film/foil caps, making them unsuitable for unregulated work, but it was merely a mention in one post without anything to back it up, so I doubt its validity a bit until I get more evidence.

 

[puppyslugg]

http://www.hagtech.com/pdf/snubber.pdf

 

[jeffreyj]

Quote:

Originally posted by Squalish ... The only thing I've heard against using those as snubbers is someone claiming that spikes in the AC damage film/foil caps, making them unsuitable for unregulated work, but it was merely a mention in one post without anything to back it up, so I doubt its validity a bit until I get more evidence.

It is better to use metallized film rather than film and foil wherever high transients are expected because a metallized dielectric is self-healing; a spike that punches through the dielectric will easily vaporize the thin layer of metallization surrounding the hole.

On the secondary side of a transformer, though, it is unlikely that spikes will ever make it through (the distributed capacitance between the windings) to cause damage, so use film-and-foil with a clear conscience (any spikes that do make it through and which are capable of destroying the snubber capacitor, will also destroy everything else down the line, so why worry?}

 

[Arzela]

Thanks all. Another question, though:
is the ~500 nf value appropriate?

 

[Voodoochile]

That is a popular value, and it was worked well for me. I have also used .01uf. I cannot offer any scientific reasoning for this, but they are the two I see spec'd most often. Probably the .01uF moreso than 500nF, come to think of it.

I use the metallized polypropylene BC/Philips caps. Wimas should be a nice little upgrade, but harder to source than the BC/Philips. I doubt it would be a worthwhile upgrade at the bridge, though. (in terms of having to buy many more than you need, perhaps)

Edit: The Hagerman article is quite informative (puppyslugg posted above). Tune it right up. I know you want to.

 

[jeffreyj]

Quote:

Originally posted by Arzela Thanks all. Another question, though: is the ~500 nf value appropriate?

Do you have an LCR bridge so you measure the leakage inductance of transformer? If not, then the proper value of snubber capacitor can't be calculated. However, you'll surely be pleased to know that most engineers determine the snubber capacitance (and the need for snubber capacitance!) empirically.

I should also state that the use of snubbers in low voltage 50/60Hz supplies is a questionable practice. If you truly feel that the reverse recovery snap is imperiling your amplifier, use Schottkys and/or go with a choke-input filter. Besides the fact that 60Hz sine waves are less than challenging to even the most lethargic rectifier, adding a snubber can actually increase the amount of noise conducted through the bridge: after all, you are adding capacitance (w/ series resistance, of course) in parallel with the junction capacitance the rectifier exhibits when it is reverse biased.